Statistics are often used to enable managers and operators to assess the behavior or performance of automated systems of many types. This application is particularly concerned with contact centers and self-service systems although the invention itself is not limited to those types of system; any automated system which can be represented using a state/event model can be used.
Previously, statistical information about the performance of contact centers and self-service systems has been generated using hard-coded methods incorporated into the processing of the contact center or self-service system itself (or an associated reporting application). The resulting statistical values calculated using the hard coded processes are then typically written out to a database from which reports are automatically generated for use by system operators. Examples of such statistics for a call center, include average call time and total number of calls answered in a specified time frame.
Such known statistic reporting methods are subject to various drawbacks. For example, operators are unable to easily obtain values of new statistics. In order to do so it would be necessary to modify and add to the hard coded process for statistic generation within the automated system itself. Thus operators are unable to obtain values of new statistics on an ad-hoc basis. Also, it is not easy or straightforward for operators to obtain information about how a particular statistical value has been obtained. For example, a statistical value such as an average may be provided. The operator is unable to quickly assess which values have been used to calculate this average without resorting to analysis of the hard-coded methods used. Another disadvantage is that because the methods for statistics generation are hard coded, then if changes are made to the operation of the automated system itself, the statistics will not necessarily remain an accurate description of the behavior of that automated system. For example, if the methods used within a contact center change slightly, the statistics generated may no longer be a good indication of the behavior of that contact center. In some cases such a change in methods used can lead to corruption of the statistical data generated.
An object of the present invention is to provide a method and system for generating statistics about the behavior of an automated system which overcomes or at least mitigates one or more of the problems mentioned above.